U.S. patent number 8,405,613 [Application Number 11/424,764] was granted by the patent office on 2013-03-26 for optimization of statistical movement measurement for optical mouse, with particular application to laser-illuminated surfaces.
This patent grant is currently assigned to EM Microelectronic-Marin SA. The grantee listed for this patent is Gil Afriat, Lawrence Bieber, Kevin Scott Buescher, James Harlod Lauffenburger, Michel Willemin. Invention is credited to Gil Afriat, Lawrence Bieber, Kevin Scott Buescher, James Harlod Lauffenburger, Michel Willemin.
United States Patent |
8,405,613 |
Bieber , et al. |
March 26, 2013 |
Optimization of statistical movement measurement for optical mouse,
with particular application to laser-illuminated surfaces
Abstract
A method for measuring relative motion between an illuminated
portion of a surface and an optical sensing device comprising a
coherent light source and a photodetector device comprising an
array of pixels and comparators for extracting motion features. The
method includes the steps (a)-(d). Step (a) includes illuminating
with the coherent light source the surface portion at a determined
flash rate. Step (b) includes detecting, using the array of pixels
a speckled light intensity pattern of the illuminated portion of
the surface for each flash. Step (c) includes extracting edge
direction data of two different types from the detected speckled
light intensity patterns. Step (d) includes measuring relative
motion between the optical sensing device and the illuminated
portion of the surface based on extracted edge direction data. The
step of extracting edge direction data further includes a
preliminary step of introducing a selecting factor, which promotes
detection of one type of edge direction data rather than another
type.
Inventors: |
Bieber; Lawrence (Colorado
Springs, CO), Willemin; Michel (Pr les, CH),
Afriat; Gil (Monument, CO), Lauffenburger; James Harlod
(Colorado Springs, CO), Buescher; Kevin Scott (Colorado
Springs, CO) |
Applicant: |
Name |
City |
State |
Country |
Type |
Bieber; Lawrence
Willemin; Michel
Afriat; Gil
Lauffenburger; James Harlod
Buescher; Kevin Scott |
Colorado Springs
Pr les
Monument
Colorado Springs
Colorado Springs |
CO
N/A
CO
CO
CO |
US
CH
US
US
US |
|
|
Assignee: |
EM Microelectronic-Marin SA
(Marin, CH)
|
Family
ID: |
38480573 |
Appl.
No.: |
11/424,764 |
Filed: |
June 16, 2006 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20070290121 A1 |
Dec 20, 2007 |
|
Current U.S.
Class: |
345/166; 345/175;
250/557; 250/208.1; 345/156; 250/221; 178/19.05; 250/559.2;
178/18.09 |
Current CPC
Class: |
G06F
3/0317 (20130101) |
Current International
Class: |
G06M
7/00 (20060101) |
Field of
Search: |
;345/161-167,156-158,426,520,173,184,581 ;700/83
;250/557,221,559.2,208.1 ;341/20 ;178/18.06,18.09,19.05 ;710/63
;702/64 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
Other References
OP. Judd, Diffraction From Circular and Irregular Apertures,
LA-5391-MS 1-4 (Los Alamos Scientific Laboratory 1973). cited by
applicant .
John P. Barton, Electromagnetic Field Calculations for a Sphere
Illuminated by a Higher-order Gaussian Beam. II. Far-Field
Scattering, 37 Applied Optics 3339-3344 (1998). cited by applicant
.
Darwin Palima et al., Generalized Phase Contrast Matched to
Gaussian Illumination, 15 Optics Express 11971-11977 (2007). cited
by applicant .
M.J. Wang et al., Investigation on the Scattering Characteristics
of Gaussian Beam from Two Dimensional Dielectric Rough Surfaces
Based on Kirchhoff Approximation, 4 Progress in Electromagnetics
Research B 223-235 (2008). cited by applicant .
p. 284 and p. 1155 of Webster's Ninth New Collegiate Dictionary
(1990). cited by applicant .
Parker, J.R. et al., "Thresholding Using an Illumination Model,"
Proceedings of the Second International Conference on Document
Analysis and Recognition, pp. 270-273, (1993). cited by applicant
.
"Contrast in Optical Microscopy," 6 pages, downloaded from
www.olympusmicro.com/primer/techniques/contrast.html on Feb. 24,
2011. cited by applicant .
Smith, Warren J. "Modern optical engineering," pp. 163-165, (2000).
cited by applicant .
Miller, John, Surf Progress Report One: Non-Gaussian Beams for
Interferometric Gravitational Wave Detectors (2005), filed herewith
as Exhibit B1. cited by applicant.
|
Primary Examiner: Lao; Lun-Yi
Assistant Examiner: Merkoulova; Olga
Attorney, Agent or Firm: Griffin & Szipl, P.C.
Claims
What is claimed is:
1. A method for measuring relative motion between a surface and an
optical sensing device, wherein the optical sensing device
comprises: (i) a coherent light source; (ii) a photodetector
device; and (iii) a motion detection processing circuit, wherein
the photodetector device comprises: (1) an array of pixels; and (2)
a plurality of comparators for extracting edge direction data,
wherein the method comprises the steps of: (A) illuminating a
portion of the surface at a determined flash rate using the
coherent light source to produce light at the determined flash
rate; (B) detecting a speckled light intensity pattern of the
illuminated portion of the surface for each flash using the array
of pixels; (C) extracting edge direction data of two different
types from the detected speckled light intensity patterns by
comparing light intensity between pixels for each light intensity
pattern, wherein the two different types of extracted edge
direction data include first edge direction data and second edge
direction data; (D) determining a measurement of the relative
motion between the optical sensing device and the surface based on
the extracted edge direction data; wherein step (C) comprises
adding, in the light intensity comparison between pixels, an offset
in the comparison of light intensity for at least one region of
light intensity patterns that promotes, for each region among the
at least one region of the light intensity patterns, detection of
the first edge direction data rather than detection of the second
edge direction data.
2. The method according to claim 1, wherein the coherent light
source illuminates the surface with a Gaussian shaped beam and
wherein the offset is adjusted as a function of the Gaussian
roll-off band.
3. The method according to claim 1, wherein the array of pixels is
aligned along first and second orthogonal axes, wherein an
inflection is detected as one type of edge direction data followed
by another opposite type of edge direction data, and wherein the
method further comprises the steps of: (E) counting along both axes
the number of inflections for each flash; (F) averaging the
inflection count of a determined number of flashes for each axis;
(G) comparing a higher average inflection count from the first axis
and the second axis with a defined average inflection count window;
and (H) adjusting said offset so that the higher average inflection
count stays within the defined window.
4. The method according to claim 3, wherein the window is defined
by minimum and maximum average inflection count values, and wherein
when the higher inflection count is greater than the maximum
average inflection count value, then the magnitude of said offset
is increased; and when the higher average inflection count is
smaller than the minimum average inflection count, then the
magnitude of said offset is decreased.
5. The method according to claim 4, wherein a targeted absolute
minimum inflection count is defined and wherein the magnitude of
the offset is increased when the higher average inflection count is
greater than the maximum average inflection count and when the
lower average inflection count is greater than the absolute minimum
inflection count.
6. A method for measuring relative motion between a surface and an
optical sensing device, wherein the optical sensing device
comprises: (i) a coherent light source; (ii) a photodetector
device; and (iii) a motion detection processing circuit, wherein
the photodetector device comprises: (1) an array of pixels; and (2)
a plurality of comparators for extracting edge direction data,
wherein the method comprises the steps of: (A) illuminating a
portion of the surface at a determined flash rate using the
coherent light source to produce light at the determined flash
rate; (B) detecting a speckled light intensity pattern of the
illuminated portion of the surface for each flash using the array
of pixels; (C) extracting edge direction data of two different
types from the detected speckled light intensity patterns by
comparing light intensity between pixels for each light intensity
pattern, wherein the two different types of extracted edge
direction data include first edge direction data and second edge
direction data; (D) determining a measurement of the relative
motion between the optical sensing device and the surface based on
the extracted edge direction data; wherein the coherent light
source provides a light intensity substantially having a maximum
value in a central region of the light source and gradually
decreasing toward peripheries from the central region of the light
source, wherein the array of pixels is aligned along first and
second orthogonal axes, and is divided into four quadrants
comprising upper left, upper right, lower left and lower right
quadrants, wherein along the first axis, when the first axis is
oriented up-down, a positive offset is added to comparators of said
plurality of comparators which are associated with both upper
quadrants, and a negative offset is added to comparators of said
plurality of comparators which are associated with both lower
quadrants; and wherein along the second axis, when the second axis
is oriented left-right, a positive offset is added to comparators
of said plurality of comparators which are associated with both
left quadrants, and a negative offset is added to comparators of
said plurality of comparators which are associated with both right
quadrants.
7. The method according to claim 6, wherein the coherent light
source illuminates the surface with a Gaussian shaped beam, wherein
a central point of the Gaussian shaped beam impinging on the array
of pixels is determined, and wherein the four quadrants are
determined in relation to the central point of the Gaussian shaped
beam.
8. The method according to claim 6, wherein an inflection is
detected as one type of edge direction data followed by another
opposite type of edge direction data, wherein the method further
comprises the steps of: (E) counting along both axes the number of
inflections for each flash; (F) averaging the inflection count of a
determined number of flashes for each axis; (G) comparing the
higher average inflection count from the first axis and the second
axis with a defined average inflection count window; and (H)
adjusting the positive and negative offsets of the comparators so
that the higher average inflection count stays within the defined
average inflection count window.
9. The method according to claim 8, wherein the window is defined
by minimum and maximum average inflection count values, and wherein
when the higher inflection count is greater than the maximum
average inflection count value, then the magnitudes of the positive
and negative offsets are increased; and when the higher average
inflection count is smaller than the minimum average inflection
count, then the magnitudes of the positive and negative offsets are
decreased.
10. The method according to claim 9, wherein a targeted absolute
minimum inflection count is defined and wherein the magnitudes of
the positive and negative offsets are increased when the higher
average inflection count is greater than the maximum average
inflection count and when the lower average inflection count is
greater than the absolute minimum inflection count.
Description
FIELD OF THE INVENTION
The present invention generally relates to pointing devices, in
particular for controlling the position of a cursor on a screen,
such as the display of a personal computer, workstation or other
computing devices having a graphic user interface. Such pointing
devices may for instance include mice, trackballs and other
computer peripherals for controlling the position of a cursor on a
display screen.
The present invention more particularly relates to the field of
optical pointing devices which comprise an optical motion sensing
device including a photodetector array for measuring the varying
intensity pattern of a portion of a surface which is illuminated
with radiation of a laser illuminated source and for extracting
information about the relative motion between the photodetector
array and the illuminated portion of the surface.
BACKGROUND OF THE INVENTION
Optical pointing devices incorporating a laser illuminated source
are already known in the art. Such laser illumination allows
optical pointing devices such as mice to work on a much wider
variety of surfaces. However the coherent nature of the
illumination results in a received image that contains generally
high spatial frequencies especially compared to viewing the same
surface with a non coherent LED illumination. This high frequency
content leads to spatial aliasing due to beyond under-sampled (in
the spatial domain). This aliasing leads to several bad effects,
such as loss of resolution or apparent "reverse" motion (the
"wagon-wheel in motion pictures" effect). One alternative to deal
with the higher spatial frequency content is to create smaller
pixels (higher spatial capability in the imager). But, two major
problems arise with smaller pixels that are a lower
mouse-speed/acceleration capability and a lower sensitivity to
light (less collection area in the pixel).
U.S. patent application Ser. No. 11/165,537, filed in the name of
the same Assignee and which is incorporated in its entirety herein
by reference, for instance discloses a method for measuring
relative motion between an illuminated portion of a surface and an
optical sensing device comprising a coherent light source and a
photodetector array, the method comprising the steps of:
illuminating under a determined gradient by means of the coherent
light source the surface portion at a determined flash rate;
detecting by means of the photodetector array speckled light
intensity pattern of the illuminated portion of the surface for a
first flash; detecting a second speckled light intensity pattern of
the illuminated portion of the surface for a second flash;
extracting motion features of two different types from the detected
first and second speckled light intensity patterns; keeping only
pairs of neighbouring motion features including one motion feature
of both different types; determining a measurement of the relative
motion between the optical sensing device and the illuminated
surface portion based on a comparison of kept motion features.
Although such a solution as disclosed in U.S. patent application
Ser. No. 11/165,537 presents several advantages in dealing with a
coherent light source in an optical pointing device, since laser
illumination has a very large spectral content, there is still a
need to further and better control the quantity of data needed to
determine a measurement of relative between the optical pointing
device and the illuminated surface portion.
SUMMARY OF THE INVENTION
One goal of the present invention is thus to implement a simpler
and more reliable method for measuring relative motion between an
illuminated portion of a surface and an optical motion sensing
device comprising a coherent light source and a photodetector
device comprising an array of pixels and comparators for extracting
motion features. For that purpose, the method comprises the steps
of:
a) illuminating by the coherent light source the surface portion at
a determined flash rate;
b) detecting by the array of pixels a speckled light intensity
pattern of the illuminated portion of the surface for each
flash;
c) extracting edge direction data of two different types from the
detected speckled light intensity patterns by comparing light
intensity between pixels;
d) determining a measurement of the relative motion between the
optical sensing device and the illuminated portion of the surface
based on extracted edge direction data; wherein the extracting edge
direction data step comprises a preliminary step consisting of
introducing a selecting factor which promotes detection of one type
of edge direction data rather than the other type.
Since an inflection is detected as one type of edge followed by the
opposite type, increasing the amount of edges of one type with
respect to the other will lead to a decrease in the overall
inflection count, and thus the density of inflections. It is
preferably proposed to introduce an adjustable offset into the
comparators in order to decrease the inflection count in the image
if it exceeds a given value. There is a need to keep the total
inflection density low enough to avoid "confusing" the inflection
tracking logic that determines movement. The added offset to these
edge detection comparators is adjusted to cause them to prefer
"one" type of edge (positive/negative) over the other type.
Thus, in accordance with a first embodiment of the invention, a
method is provided for measuring relative motion between an
illuminated portion of a surface and an optical sensing device
comprising a coherent light source and a photodetector device
comprising an array of pixels and comparators for extracting motion
features, wherein the method comprises the steps of: (a)
illuminating by means of the coherent light source the surface
portion at a determined flash rate; (b) detecting by means of the
array of pixels a speckled light intensity pattern of the
illuminated portion of the surface for each flash; (c) extracting
edge direction data of two different types from the detected
speckled light intensity patterns by comparing light intensity
between pixels; (d) determining a measurement of the relative
motion between the optical sensing device and the illuminated
portion of the surface based on extracted edge direction data;
wherein the extracting edge direction data step comprises a
preliminary step consisting of introducing a selecting factor that
promotes detection of one type of edge direction data rather than
the other type. Furthermore, other advantageous embodiments of the
invention are summarized as follows.
In accordance with a second embodiment of the present invention,
the first embodiment is modified so that the array of pixels is
aligned along first and second axes and divided into four quadrants
including upper left, upper right, lower left and lower right
quadrants, and wherein the selecting factor is introduced by adding
to comparators along the first axis: (i) a positive offset for both
left quadrants; (ii) a negative offset for both right quadrants;
and by adding to comparators along the second axis: (i) a positive
offset for both lower quadrants; and (ii) a negative offset for
both upper quadrants. In accordance with a third embodiment of the
present invention, the second embodiment is modified so that it
further comprises a step consisting of determining the central
point of a Gaussian illumination of the coherent light source and
wherein the four quadrants are determined in relation with the
central point of the Gaussian illumination.
In accordance with a fourth embodiment of the present invention,
the first embodiment is modified so that the coherent light source
illuminates with a Gaussian shaped beam and wherein the selecting
factor is introduced by adding to the comparators an offset that is
adjusted in function of the Gaussian roll-off band. In accordance
with a fifth embodiment of the present invention, the fourth
embodiment is further modified so that the offset is inverted
according to the Gaussian roll-off sign.
In accordance with a sixth embodiment of the present invention, the
second embodiment is further modified so that an inflection is
detected as one type of edge followed by the opposite type, wherein
the method further comprises the steps of: (e) counting along both
axes the number of inflections for each flash; (f) averaging the
inflection count of a determined number of flashes for each axes;
(g) comparing the higher inflection count of either the first or
the second axis with a defined average inflection count window; and
(h) adjusting the offset voltage so that the count will stay within
the defined window. In accordance with a seventh embodiment of the
present invention, the sixth embodiment is further modified so that
the window is defined by minimum and maximum average inflection
count values and wherein when the higher inflection count is
greater than the maximum average inflection count value, then the
offset voltage is increased, and when the higher average inflection
count is smaller than the minimum average inflection count, then
the offset voltage is decreased. In accordance with an eighth
embodiment of the present invention, the seventh embodiment is
further modified so that a targeted absolute minimum inflection
count is defined and wherein the offset is increased if the higher
average inflection count is greater than the maximum average
inflection count and if the lower average inflection count is
greater than the absolute minimum inflection count.
A refinement can be made to mitigate one issue with the simplest
implementation of this offset. That issue is due to the non-uniform
illumination of the optical source. Generally the laser will
illuminate with a "Gaussian" shaped beam, and so the image will
naturally have a Gaussian roll-off convolved with the actual image
from the surface. The edges of the roll-off confuse the offset
algorithm, and so the action in those regions is not what is
desired. On one edge, the number of inflections is decreased, and
on the other edge the number is increased. To get around this, the
offset is advantageously dependent on which image quadrant the
edge-comparators are in.
For that purpose, the photodetector array of pixels is aligned
along first and second axes and divided into four quadrants upper
left, upper right, lower left and lower right and the selecting
factor is introduced by adding to comparators along the first axis
a positive offset for both left quadrants and a negative offset for
both right quadrants, and by adding to comparators along the second
axis a positive offset for both lower quadrants and a negative
offset for both upper quadrants.
A further refinement can be made to deal with a further problem.
The optical alignment of the laser illumination might not be
perfect, and so the central point of the Gaussian illumination
might not correspond to the central point of the pixel array. To
handle this issue, the algorithm measures the central point of
illumination (with respect to X and Y), and changes the sign of the
offset polarity of each edge comparator to effectively move the
matrix central point to correspond to the illumination central
point.
For that purpose, the method further comprises a step consisting of
determining the central point of a Gaussian illumination of the
coherent light source and wherein the four quadrants are determined
in relation with the central point of the Gaussian
illumination.
Alternatively, since the coherent light source illuminates with a
Gaussian shaped beam, the selecting factor is introduced by adding
to the comparators an offset being adjusted in function of the
Gaussian roll-off band. It may be further provided that the offset
is inverted according to the Gaussian roll-off sign.
Since the algorithm reads the inflection count (along both axes X
and Y) during every "flash" of the laser. A number of flashes (N)
are averaged to produce an average inflection count (for X and Y),
and the larger of the X or Y count is checked against a defined
window. Then, the algorithm adjusts the offset voltage so that the
count will stay within the defined window.
Additional enhancements are included to better handle various "end"
conditions. For example, while the highest count of X or Y controls
the algorithm, the other lower count of X or Y is also checked to
make sure that it does not drop below a certain threshold. If it
does, the offset may not be adjusted further.
BRIEF DESCRIPTION OF THE DRAWINGS
Other aspects, features and advantages of the present invention
will be apparent upon reading the following detailed description of
non-limiting examples and embodiments made with reference to the
accompanying drawings.
FIG. 1 represents a generalized schematic bloc diagram of an
optical pointing device;
FIG. 2 shows a three-dimensional image of the laser illumination
which has a fairly steep Gaussian gradient;
FIG. 2a represents a cut section of the plan A-A in FIG. 2;
FIG. 2b represents a photodetector array cut into four
quadrants;
FIG. 3 is a schematic illustration of edge inflection conditions,
or peaks and nulls, derived from a sequence of edge direction
conditions along a determined axis of the photodetector array.
DETAILED DESCRIPTION OF THE INVENTION
The following description, which concerns a method for measuring
relative motion between an illuminated portion of a surface and an
optical sensing device comprising a coherent light source and a
photodetector comprising an array of pixels and comparators for
extracting motion features, is given by way of a non limiting
example in relation with FIGS. 1 to 3.
Algorithms that may be used for determining the measurement of the
relative motion between the optical sensing device and the
illuminated portion of the surface based on extracted edge
direction data are given by way of example in US Patent Application
Publication No 2005/0062720 filed in the name of the same Assignee
and enclosed herewith in its entirety by way of reference. It is
understood that various adaptations may be done on these
algorithms.
According to these algorithms, extracted motion features are
defined as edge direction data given by comparing light intensity
between pixels, namely a first edge condition, or positive edge,
defined as a condition wherein the light intensity of a first pixel
is less than the light intensity of a second pixel, and a second
edge condition, or negative edge, defined as a condition wherein
the light intensity of the first pixel is greater than the light
intensity of the second pixel.
Further, referring to the "Peak/Null Motion Detection" algorithm
each row and column of the photodetector array are further analysed
to find specific inflection conditions (hereinafter defined as a
first inflection condition, or "peak", and a second inflection
condition, or "null") in the direction of successive edges along a
selected axis (in practice along both the X and Y axes). As
illustrated in FIG. 3, the first inflection condition, or peak, is
defined as the succession, along a determined axis (X or Y), of a
positive edge (arrow pointing upwards in FIG. 3) followed by a
negative edge (arrow pointing downwards in FIG. 3). Similarly, the
second inflection condition, or null, is defined as the succession,
along the determined axis, of a negative edge followed by a
positive edge.
Considering now FIG. 1, it represents a generalized schematic bloc
diagram of an optical pointing device. It comprises a photodetector
array 100 including a plurality of pixels, this photodetector array
100 being coupled to processing means 110 (or motion detection
processing circuit) for processing the signals outputted by the
photodetector array 100.
A comparator array 120 may be interposed between processing means
110 and array 100, this comparator array 120 including a plurality
of comparator circuits each for comparing the light intensity of a
first pixel of array 100 with the light intensity of a second pixel
of array 100 and for outputting resulting motion feature
conditions.
The optical pointing device further comprises at least one coherent
light source 130 such as a laser illumination source, which
produces radiation at a determined flash rate, that impinges with a
determined gradient on a portion of a surface S. Surface S may be a
planar or non-planar surface, such as a surface over which the
pointing device is moved (as in the case of an optical mouse), the
surface of a ball (as in the case of an optical trackball) or any
other suitable surface that may provide an appropriate speckled
intensity pattern for detection by photodetector array 100.
Processing means 110 is further adapted to communicate in a
bi-directional manner with an interface 140 that communicates in
turn with a host system (not illustrated) over a bus 150. Cursor
control signals (and eventually other signals related to the
optical pointing device) are supplied to the host system over bus
150. Processing means 110 may also receive information, such as
configuration signals, over bus 150 from the host system.
Processing means 110 is essentially designed to intermittently
sample the pixel outputs of photodetector array 100 in accordance
with a defined sequence. The information of two successive samples
or speckled images is compared and a relative motion measurement is
extracted by processing means 110. The adequate cursor control
signals are then derived from the relative motion measurement and
transmitted to the host system via line interface 140.
However, as it has been already mentioned in the introduction of
the specification, the light intensity pattern detected by
photodetector device 100 forms a speckled image which presents to
many motion features which render motion detection less reliable.
For that purpose according to the present invention, processing
means 110 are provided with a very simple selecting factor 160
which promotes detection of one type of motion feature rather than
the other type. Such simple selecting factor will be explained and
better understood below in relation with FIGS. 2, 2a and 2b.
FIG. 2 shows a three-dimensional image of the laser illumination
which has a fairly steep Gaussian gradient. The plan defined by
axes X and Y represents the photodetector array and axis Z
represents the detected light intensity. As can be easily seen on
this image, the light intensity increases or decreases according to
the quadrant and axis considered. Thus if one cuts the array of
pixels in four quadrants (as shown on FIG. 2b), namely, the upper
left, upper right, lower left and lower right quadrants, the light
intensity (LI) departing from edges to the centre of the array of
pixels will change in the following manner:
in the upper left quadrant, the light intensity increases along
both X and Y axes;
in the upper right quadrant, the light intensity increases along X
axis and decreases along axis Y;
in the lower left quadrant, the light intensity decreases along the
X axis and increases along the Y axis; and
in the lower right quadrant, the light intensity decreases along
both X and Y axes.
As mentioned previously, the purpose of the algorithm is to control
the average inflection count that the sensor sees. Since laser
illumination has a very large spectral content, it leads to a dense
inflections image, which causes the motion detection algorithms of
the sensor to alias. The aliasing is usually seen as degradation of
resolution with speed and/or acceleration detection
(loss-of-tracking event) failure. By controlling the average
inflection count, one also controls indirectly the density of the
inflections image and aliasing may be prevented.
By adding an adjustable offset to the edge detection comparators in
function of the quadrant, the inflection count is decreased.
Actually, adding an offset to the edge detection comparators causes
them to "prefer" one type of edges (positive/negative) over the
other. Since an inflection is detected as one type of edge followed
by the opposite type, increasing the amount of edges of one type on
the account of the other will necessarily lead to a decrease in the
inflection count.
For that purpose, according to an embodiment of the present
invention and in accordance with the quadrant considered, the
offset is adjusted in the following manner:
a negative offset for both lower quadrants along the X axis;
a positive offset for both upper quadrants along the X axis;
a positive offset for both left quadrants along the Y axis; and
a negative offset for both right quadrants along the Y axis.
Since the central point of the Gaussian illumination of the
coherent light source is not always centered with respect to the
array of pixels, the four quadrants are preferably determined in
relation with the actual central point of the Gaussian
illumination.
According to another embodiment of the present invention, since the
coherent light source illuminates with a Gaussian shaped beam, the
selecting factor is introduced by adding to the edge comparators an
offset being adjusted in function of the Gaussian roll-off band.
Preferably the offset is inverted according to the Gaussian
roll-off sign.
Advantageously, the algorithm reads the X and Y inflection count
for every flash. N flashes are averaged to produce X and Y average
inflection count. For instance, the number of consecutive flashes
used to calculate the average inflection count may by default be
fixed to 4. The algorithm adjusts the offset voltage, so the
largest between X and Y average inflection count falls within a
pre-defined window of inflection count. The window is defined for
both axis X and Y, with targeted maximum and minimum average
inflection counts, respectively NinfAveHi and NinfAveLo.
The offset control algorithm will adjust the comparators offset so
to the higher average inflection count (the higher between X
average inflection count and Y average inflection count) will fall
within the window that is defined by minimum and maximum values
NinfAveLo and NinfAveHi. Thus if the higher inflection count is
larger than the maximum average inflection count NinfAveHi, the
offset voltage will be increased. If the higher average inflection
count is smaller than the minimum average inflection count
NinfAveLo, the offset voltage will be decreased.
Since the algorithm is looking at the higher average inflection
count, a precaution should be taken regarding the lower inflection
count. One would not want this inflection count to go below some
level (an extremely low inflection count also leads to performance
degradation). Thus, the algorithm allows an increase of the offset
voltage if the higher average inflection count is larger than the
maximum average inflection count NinfAveHi and if the lower average
inflection count is higher than a targeted absolute minimum
inflection count NinfAveMin.
FIG. 2a represents a cut section of the plan A-A in FIG. 2. Adding
a positive offset when the light intensity increases has the effect
to accentuate the light intensity increase due to the Gaussian
shaped beam and eliminate succession of positive and negative
edges, i.e. inflection, due to small intensity differences and
therefore reduce the total number of the inflection count.
* * * * *
References